Osmotic stress responses of individual white oak (Quercus section, Quercus subgenus) genotypes cultured in vitro

Our reference: JPLPH 51807 P-authorquery-v9 AUTHOR QUERY FORM Journal: JPLPH Please e-mail or fax your responses and any corrections to: E-mail: [email protected] Article Number: 51807 Fax: +353 6170 9272 Dear Author, Please check your proof carefully and mark all corrections at the appropriate place in the proof (e.g., by using on-screen annotation in the PDF file) or compile them in a separate list. Note: if you opt to annotate the file with software other than Adobe Reader then please also highlight the appropriate place in the PDF file. To ensure fast publication of your paper please return your corrections within 48 hours. For correction or revision of any artwork, please consult http://www.elsevier.com/artworkinstructions. Any queries or remarks that have arisen during the processing of your manuscript are listed below and highlighted by flags in the proof. Click on the ‘Q’ link to go to the location in the proof. Location in Query / Remark: click on the Q link to go article Please insert your reply or correction at the corresponding line in the proof The reference given here is cited in the text but is missing from the reference list – please make the list complete or remove the reference from the text: ‘Johnson et al. (2002)’. Q1 Please confirm that given names and surnames have been identified correctly. Q2 Reference‘Johnsonetal.(2002)’iscitedinthetextbutnotprovided in the reference list. Please provide it in the reference list or delete the citation from the text. Please check this box or indicate your approval if you have no corrections to make to the PDF file Thank you for your assistance. Please cite this article in press as: Demeter Z, et al. Osmotic stress responses of individual white oak (Quercus section, Quercus subgenus) genotypes cultured in vitro. J Plant Physiol (2013), http://dx.doi.org/10.1016/j.jplph.2013.09.013 ARTICLE IN PRESS G Model JPLPH 51807 1–9 Journal of Plant Physiology xxx (2013) xxx– xxx Contents lists available at ScienceDirect Journal of Plant Physiology j o ur nal homepage: www.elsevier.com/locate/jplph Physiology1 Osmotic stress responses of individual white oak (Quercus section, Quercus subgenus) genotypes cultured in vitro 2 3 Zita Demetera, Péter Kanalasa, Csaba Máthéa,∗, Klára Csekeb, Erzsébet Sz˝ oll˝ osia, Q1 Márta M-Hamvasa, Katalin Jámbrika, Zoltán Kissa, Ilona Mészárosa,∗ 4 5 aUniversity of Debrecen, Faculty of Science and Technology, Department of Botany, PO Box 14, H-4010 Debrecen, Hungary6 bHungarian Forest Research Institute, PO Box 30/A, H-9600 Sárvár, Hungary7 8 a r t i c l e i n f o 9 10 Article history:11 Received 25 June 201312 Received in revised form 11 September 2013 13 14 Accepted 22 September 201315 Available online xxx 16 Keywords:17 Quercus tissue culture18 Osmotic stress19 Recovery20 Guaiacol peroxidase 21 Single-strand preferring nuclease22 a b s t r a c t White oaks (Quercus section, Quercus subgenus) are widely distributed in Europe. Quercus petraea (sessile oak), an economically important species is predicted to be affected by climate change. Q. pubescens (pubescent oak) and Q. virgiliana (Italian pubescent oak) are economically less important, drought tolerant species. Frequent hybridization of white oaks was observed and currently the introgression of Q. pubescens and Q. virgiliana in non-mediterranean regions of Europe has been reported. Our goal was to use tissue cultures established from individual trees of the above taxa and their putative hybrids, all present in the forest stand of Síkf˝ okút LTER Research Area (NE Hungary) as simple experimental model systems for studying drought/osmotic stress tolerance. Tissue cultures are more suitable models for such studies, than seedlings, because they are genetically identical to the parent plants. Polyethylene glycol (PEG6000) treatments were used for this purpose. The identification of taxa was based on leaf morphological traits and microsatellite analysis and showed that Q. petraea is genetically distinct to all other taxa examined. We established six callus lines of Quercus. As expected, in Q. petraea cultures PEG6000 induced severe loss of fresh weight and the ability to recover after removal of the osmoticum, which was not characteristic for Q. pubescens and Q. virgiliana. Putative hybrids exhibited an intermediate response to osmotic stress. Activity gels showed the increase of single-strand preferring (SSP) nuclease and no significant change of guaiacol-peroxidase activities in drought-sensitive genotypes/cultures and no significant increase of SSP nuclease activities accompanied with increases of guaiacol-peroxidase activities in drought-tolerant ones. This indicates that drought/osmotic stress tolerance is associated to increased capacity of scavenging reactive oxygen species and hence less susceptibility to DNA damage. Our results confirm that tissue cultures of oak are suitable model systems for studying drought/osmotic stress responses. © 2013 Published by Elsevier GmbH. Introduction 23 Oaks (Quercus spp.) are widespread in Europe and play an24 important ecological and silvicultural role. High genetic and mor-25 phological variability has been reported for the Quercus section 26 (=Lepidobalanus; or white oaks sensu Nixon, 1993) of the Quer-27 cus subgenus (Herzog, 1996; Gailing et al., 2007). The three main28 white oak species are Quercus robur L., Q. petraea (Matt.) Liebl and29 Q. pubescens Willd. Several other oak taxa have also been dis30 tinguished from the European broadleaved forests, of which Q.31 dalechampii Ten., Quercus polycarpa Schur, and Q. virgiliana Ten.32 are important. The two former ones resemble to Q. petraea sensu33 stricto and are usually included in the aggregate of Q. petraea sensu34 ∗Corresponding author. Tel.: +36 52512900; fax: +36 52512943. E-mail addresses: [email protected] (C. Máthé), [email protected] (I. Mészáros). lato, while Q. virgiliana belongs to the aggregate of Q. pubescens 35 sensu lato (Schwarz, 1936a,b; Bordács et al., 2002). Interspecific 36 hybridization is very common between white oak species (Curtu 37 et al., 2007; Lepais et al., 2009; Salvini et al., 2009; Lepais and 38 Gerber, 2011) which increases the genetic diversity in natural pop39 ulations (Borovics et al., 1998; Gömöry and Schmidtová, 2007; 40 Kanalas et al., 2008). 41 European white oaks differ in the preference for ecological con42 ditions and grow in various habitats. The distribution of oaks is 43 mostly dependent on their capacity to resist drought or excess of 44 water in the soil or even the two phenomena successively (Jones, 45 1959; Johnson et al., 2002). Among the three main white oak species Q246 Q. pubescens is the most drought tolerant one and occupies warm 47 and xeric sites in Europe (Borovics et al., 1998; Yurukov and Zhelev, 48 2001; Thomas et al., 2002; Gallé et al., 2007; Siam et al., 2009). Q. 49 petraea grows predominantly on mesic or relatively dry sites on 50 lower altitude slopes and ridges, whereas Q. robur can populate 51 lowland sites with wet and temporarily waterlogged soils (Jones, 52 0176-1617/$ – see front matter © 2013 Published by Elsevier GmbH. http://dx.doi.org/10.1016/j.jplph.2013.09.013 Please cite this article in press as: Demeter Z, et al. Osmotic stress responses of individual white oak (Quercus section, Quercus subgenus) genotypes cultured in vitro. J Plant Physiol (2013), http://dx.doi.org/10.1016/j.jplph.2013.09.013 ARTICLE IN PRESS G Model JPLPH 51807 1–9 2Z. Demeter et al. / Journal of Plant Physiology xxx (2013) xxx– xxx Table 1 Identity of Quercus taxa used in this study, based on 16 leaf quantitative traits measured according to Borovics (2000) and Kanalas et al. (2008, 2009). Explant code number Identity A149 Q. petraea D137 Q. petraea × Q. dalechampii D89 Q. petraea × Q. pubescens A211 Q. virgiliana × Q. polycarpa B50 Q. virgiliana A75 Q. pubescens 1959; Aas, 1998). Q. dalechampii Ten. (Theodoropoulos et al., 1995)53 and Quercus polycarpa Schur (Matula, 2009) are distinct from Q.54 petraea sensu stricto in ecological requirements. Both are more55 drought tolerant and grow in warmer sites. Q. dalechampii Ten.56 and Q. virgiliana have been described from arid sites of southern57 Europe (Ofletea et al., 2011). Those oak species or their hybrids58 that are more capable to stand dry and hot summer periods could59 be important tools for the future forestry as global warming and60 frequent drought events are predicted to increase the pressures to61 European oak forests (IPCC, 2007). 62 In Hungary mixed forests of sessile oak and Turkey oak (Querce-63 tum petraeae-cerris) cover the largest part of forested area. The64 Síkf˝ okút Long-term Ecological Research Site (LTER) (Bükk Moun-65 tains, north-eastern Hungary) represents this forest type (Jakucs,66 1985) and has international reputation as a former IBP and MAB67 area and current LTER Europe network member. The area is situated68 in the transition between forest and forest-steppe zone (47◦55N,69 20◦26E, 320–340 m a.s.l.) and is vulnerable to the climate change70 (Mészáros et al., 2007). Intense forest monitoring research has been71 running for 40 years in the site. The forest stand is currently charac72 terized by the dominance of taxa from Quercus section and Q. cerris73 (Mészáros et al., 2007; Kanalas et al., 2008, 2009). Based on leaf74 morphological traits the assignment of trees to white oak species75 showed that this forest stand is mostly composed of Q. petraea76 senso stricto (70.2%). 1% of sampled trees was assigned to Q. poly-77 carpa, 2.5% was assigned to Q. virgiliana and 4% was assigned to78 Q. pubescens. 22, 3% of trees were considered as putative hybrids:79 Q. petraea × Q. dalechampii (5.1%), Q. petraea × polycarpa (2.5%), Q.80 petraea × Q. pubescens (5.6%), Q. petraea × Q. virgiliana (7.1%) and81 Q. virgiliana × Q. pubescens (2%) (Kanalas et al., 2008, 2009). The82 percentage of hybrids is relatively high as compared to oak commu-83 nities from other European regions (see Curtu et al., 2007; Gugerli84 et al., 2007 for examples). The oak decline observed in the 1980s85 in Europe approached this site too, the die-back of trees occurred86 primarily in population of Quercus petraea sensu lato. 87 Plant tissue culture techniques including callus and cell sus-88 pension cultures offer many advantages – e.g. they provide fully89 controllable systems – for physiological/biochemical studies. Con-90 sequently they were applied for oak species as well, e.g. for studying91 the role of ABA in the maturation of Q. ilex embryos (Mauri and92 Manzanera, 2004) or of dehydrin proteins in Q. robur somatic93 embryos (ˇ Sunderlíková et al., 2009). Even though in vitro cultures94 represent different developmental stages and gene expression pat-95 terns than mature plants, cell and callus cultures of sessile oak and96 pedunculate oak proved to be excellent models and are currently97 used for the study of osmotic stress-related transcriptional changes 98 and other physiological responses. They are thought to be suitable99 for physiological studies at tissue, cell and molecular level (Gleeson100 et al., 2004; Porth et al., 2005; ˇ Sunderlíková et al., 2009).101 In our study we have selected trees of several putative 102 drought-sensitive and -tolerant taxa belonging to Quercus section103 co-occurring in the forest stand of Síkf˝ okút Project LTER site (Jakucs,104 1985) for establishing in vitro cultures (Table 1). What could be the105 importance of tissue cultures in this respect? Instead of studying 106 the effect of environmental stress in mature trees we can examine107 physiological responses in in vitro cultures under controlled condi108 tions. In vitro cultures established from vegetative tissues are more 109 likely to reflect the genetic background of original plants/explants, 110 than e.g. seedlings with uncertain genetic origin. Our principal aim 111 was to study drought/osmotic stress tolerance of different oak 112 genotypes in a model – in vitro culture system. For this purpose, 113 we needed to establish a procedure for creating stable tissue cul114 tures since there were no literature data or culture ready-to-use 115 for those oak genotypes. The use of axenic cultures allowed us to 116 avoid potential superimposing effects of multiple environmental 117 conditions. 118 Peroxidases (E.C. 1.11.1.7) play a role in the scavenging of reac119 tive oxygen species (ROS) known for elevated levels during drought. 120 The presence of scavenging systems is a good indicator of drought 121 tolerance (Reddy et al., 2004). In the absence of protection against 122 oxidative stress, DNA and RNA damage occurs. This may be accom123 panied by increases in the activity of nucleases, among them, single 124 strand preferring (SSP) nucleases (EC 3.1.30.1) (Reddy et al., 2004; 125 Roldán-Arjona and Ariza, 2009). Therefore, we studied PEG6000 126 induced changes in stress enzyme – peroxidase, nuclease – activ127 ities that accompanied alterations of fresh weight and recovery of 128 oak tissue cultures. PEG6000 is widely used for modeling osmotic 129 stress in higher plants, due to minimal side-effects and the incapa130 bility of plant cells to metabolize it (see Hohl and Schopfer, 1991; 131 Guóth et al., 2010 for examples). Our basic hypothesis was that 132 osmotic stress responses will be different for in vitro cultures estab133 lished from explants derived from individuals of drought sensitive 134 (Q. petraea sensu stricto) from those of drought tolerant species (Q. 135 pubescens, Q. virgiliana). Since Q. dalechampii is closely related to Q. 136 petraea (Borovics et al., 1998), thus the sensitivity to drought of tis137 sue cultures of individuals representing hybrids between them (Q. 138 petraea × Q. dalechampii) was expected to be close to drought sen139 sitive Q. petraea. In contrast, cultures of Q. virgiliana × Q. polycarpa 140 hybrids were expected to be more tolerant to osmotic stress than Q. 141 polycarpa, due to the higher drought tolerance of Q. virgiliana par142 ent. Cultures of Q. petraea × Q. pubescens hybrid were presumably 143 intermediary between the two parents with respect to drought tol144 erance. If in vitro experiments confirm the expected responses for 145 individuals of non-hybrid taxa (i.e. drought sensitivity of Q. petraea 146 and tolerance of Q. pubescens and Q. virgiliana), they can be used for 147 testing of cultures derived from Quercus genotypes (e.g. putative 148 hybrids) with unknown drought sensitivity. 149 The main goal of this study was to offer a model system based 150 on in vitro cultures of selected individual white oak trees for the 151 estimation of their genotype-dependent drought tolerance. 152 Materials and methods 153 Plant material and tissue culture 154 Plant material was collected in the oak forest of Síkf˝ okút 155 Research Area, North-Eastern Hungary (47◦55N, 20◦26E, 156 320–340 m a.s.l.) in early spring before leaf flush. Young (2–4 157 years old) shoots were cut from the lower canopy of 105–110 158 years old mature trees of Quercus petraea (Mattuschka) Lieblein, 159 Q. pubescens Willd., Q. virgiliana Ten. and of putative hybrids 160 and transferred to laboratory. Each explant was collected from 161 a selected single tree per genotype listed in Table 1 and used 162 for establishment of tissue culture and subsequent physiological 163 experiments. All explants originated from the same ecological 164 conditions. After collection shoots were kept under conditions 165 of 22 ± 2◦C, 20 ␮mol m−2s−1PFD and 12/12 h photoperiod until 166 the development of young leaves (12 ± 4 mm length). These 167 young leaves were used as explants for the induction of tissue 168 cultures. The induction and maintenance of callus production 169 Please cite this article in press as: Demeter Z, et al. Osmotic stress responses of individual white oak (Quercus section, Quercus subgenus) genotypes cultured in vitro. J Plant Physiol (2013), http://dx.doi.org/10.1016/j.jplph.2013.09.013 ARTICLE IN PRESS G Model JPLPH 51807 1–9 Z. Demeter et al. / Journal of Plant Physiology xxx (2013) xxx– xxx 3 Table 2 Key parameters for the SSR markers used for microsatellite analysis of white oak trees used in this study. SSR primer pair Repeat motif Primer sequence No. of alleles/ no. of trees Observed heterozygosity Reference for the method of microsatellite analysis ZAG1/5 F (GT)5(GA)9 GCTTGAGAGTTGAGATTTGT 7/6 0.500 Steinkellner et al. (1997) ZAG1/5 R GCAACACCCTTTAACTACCA ZAG 9 F (AG)12 GCAATTACAGGCTAGGCTGG 7/6 0.833 Steinkellner et al. (1997) ZAG 9 R GTCTGGACCTAGCCCTCATG ZAG 110 F (AG)15 GGAGGCTTCCTTCAACCTACT 9/6 1.000 Steinkellner et al. (1997) ZAG 110 R GATCTCTTGTGTGCTGTATTT ZAG 11 F (TC)22 CCTTGAACTCGAAGGTGTCCTT 4/6 0.833 Kampfer et al. (1998) ZAG 11 R GTAGGTC A AAACCATTGGTTGACT ZAG 96 F (TC)20 CCCAGTCACATCCACTACTGTCC 7/6 1.000 Kampfer et al. (1998) ZAG 96 R GGTTGGGAAAAGGAGATCAGA ZAG 112 F (GA)32 TTCTTGCTTTGGTGCGCG 2/6 0.167 Kampfer et al. (1998) ZAG 112 R GTGGTCAGAG ACTCGGTAAGTATTC was achieved on WPM medium (Woody Plant Medium, Lloyd 170 and McCown, 1980) solidified with 0.8% (w/v) agar (Difco,171 Lawrence, KS, USA). The plant growth regulators (PGRs) used were 172 0.05–4 mg L−1(0.53–21.5 ␮M) ␣-naphthaleneacetic acid (NAA)173 and 0.05–1 mg L−1(0.25–0.5 ␮M) indole-3-butyric acid (IBA) as174 auxins and 0.1–4 mg L−1(0.44–17.7 ␮M) N6-benzyladenine (BA) as175 a cytokinin. All PGRs were from Sigma–Aldrich, Budapest, Hungary.176 The design of PGR content of tissue culture media was based on177 the methods of Seckinger et al. (1979), Tanaka et al. (1995), Cuenca178 et al. (1999) and Toribio et al. (2004). Growth conditions for in vitro 179 culture were: 14/10 h photoperiod with a photon fluence rate180 of 10 ␮mol m−2s−1during the light period and temperatures of181 22 ± 2◦C/18 ± 2◦C.182 Determination of taxonomical status of white oak individuals by 183 leaf morphological traits184 Leaf morphological traits were assessed in all trees selected for185 establishment of in vitro cultures in a former comprehensive tax-186 onomic survey of 198 white oak trees in the forest stand (Kanalas187 et al., 2008, 2009). Briefly, in this survey five leaves were collected188 from the lower canopy of trees. Altogether means of 16 leaf quan189 titative traits were used in numeric classification analysis worked190 out previously for each taxon (Borovics, 2000; Kanalas et al., 2008,191 2009). This analysis allowed to classify the individual trees as oak 192 species sensu stricto and hybrids.193 Microsatellite (SSR) analysis 194 For DNA extraction winter bud samples of the individual trees195 listed in Table 1 were used. After grinding with liquid nitrogen the196 extraction was carried out by the Qiagen Plant Mini Kit (BioMarker, 197 Gödöll˝ o, Hungary). DNA concentration of extracts was checked by198 gel electrophoresis on a 0.5% agarose (Roth Roti®garose NEEO, RK199 Tech, Budapest, Hungary) gel. Polymerase chain reactions were per-200 formed with the following SSR markers (fluorescent dyes at the 201 5-ends are indicated in brackets): ZAG 1/5 (6-FAM), ZAG 9 (6-FAM),202 ZAG 110 (HEX) (Steinkellner et al., 1997), ZAG 11 (TET), ZAG 96203 (TET), ZAG 112 (HEX) (Kampfer et al., 1998). Key parameters for204 SSR markers are presented in Table 2. PCR mastermixes and opti205 malisation procedures were made up according to Steinkellner et al.206 (1997) and Kampfer et al. (1998) and comprised the following com-207 ponents for 15 ␮L final reaction volume: 5× buffer (PromegaGoTaq208 Flexi) 4 ␮L; MgCl21 mM (0.6 ␮L) in case of markers ZAG 1/5 and209 ZAG 9, 2 mM (1.2␮L) in case of the other markers; Primer F and210 R (Roth Roti®garose NEEO, RK Tech, Budapest, Hungary): 0.25 pM211 (0.375 ␮L) each for ZAG 9, 0.75 pM (1.125 ␮L) for ZAG 1/5 and212 0.34pM (0.5 ␮L) for the remaining markers; dNTPmix (Promega213 10 mM) 0.4 ␮L; polymerase enzyme (PromegaGoTaq Flexi) 0.4 U;214 DNA sample 1 ␮L (approx. 10 ng/␮L). For the PCRs an Eppendorf 215 Mastercycler Gradient thermocycler was used with the following 216 program: in case of ZAG 1/5 and ZAG 9 initial denaturation 95◦C217 15 min, denaturation 95◦C 50 s, primer annealing 55◦C/ZAG 9 and 218 65◦C/ZAG 1/5 50 s, elongation 72◦C 1 min 45 s, repetition of last 219 three steps in 35 cycles, final elongation 72◦C 10 min. In case of the 220 other four markers: initial denaturation 95◦C 15 min, denaturation 221 95◦C 30 s, primer annealing 50◦C 30 s, elongation 65◦C 1 min 30 s, 222 repetition of last three steps in 35 cycles, final elongation 65◦C223 15 min. Fragment analyses were carried out by an ABI Prism 310 224 genetic analyser (Applied Biosystems Life Technologies, Budapest, 225 Hungary) in multiplexed runs. C matrix set and TAMRA 500 size 226 standard were applied (Applied Biosystems Life Technologies). The 227 evaluation of fragment sizes was done by the GeneMapper soft228 ware. 229 The raw genotype data set with the SSR fragment length sizes 230 was analyzed by the GenAlEx 6.4 (Peakall and Smouse, 2006) pop231 ulation genetic software. The genetic distances between pairs of 232 individuals were calculated based on the shared alelle frequen233 cies. The genetic relationship among samples was represented on a234 dendrogram constructed with the unweighted pair-group average 235 amalgamation method (UPGMA) (Sneath and Sokal, 1973). For this 236 purpose the Cluster Analyses option of Statistica 6.0 software was 237 used. 238 PEG treatments and recovery experiments 239 Callus cultures were grown on solidified WPM medium (see 240 Results section for PGR content). They were transferred on liq241 uid medium of the same composition (2 mL culture medium in 242 10 mL sterile plastic flasks, Labsystem, Budapest, Hungary) and 243 treated for 24 h with 0, 5, 10, 20 and 40% (w/v) polyethylene gly244 col 6000 (PEG6000, VWR, Leuven, Belgium). PEG6000 solutions 245 corresponded to osmotic potentials of −0.05, −0.15, −0.49 and 246 −1.75 MPa. During treatments with the osmoticum, cultures were 247 gently shaken (100 rpm) on a rotatory shaker (E. Bühler KS-15, E. 248 Bühler GmbH, Hechingen, Germany). Fresh weight (FW) of calli 249 was measured at the start and the end of PEG6000 treatments 250 by means of an analytical balance (Model AA 200 DS, accuracy 251 ±10 ␮g, Denver Instrument Co., Arvada). FW was in the range of 252 20–50 mg. Before FW measurement at the start of experiments, 253 excess liquid medium was removed by gentle centrifugation with254 out affecting callus growth and viability (1000 rpm, 5 s on a Heraeus 255 Biofuge, Kendro Laboratory Products, Harau, Germany). Enzyme 256 activities (see below) were measured directly after treatments 257 with the osmoticum. Percentage of FW increase was calculated 258 on the basis of the difference between FW at the end and at the 259 start of experiments. Recovery experiments were performed as fol260 lows: following PEG6000 treatments, cultures from PEG containing 261 Please cite this article in press as: Demeter Z, et al. Osmotic stress responses of individual white oak (Quercus section, Quercus subgenus) genotypes cultured in vitro. J Plant Physiol (2013), http://dx.doi.org/10.1016/j.jplph.2013.09.013 ARTICLE IN PRESS G Model JPLPH 51807 1–9 4Z. Demeter et al. / Journal of Plant Physiology xxx (2013) xxx– xxx medium were washed out two times by gentle shaking (100 rpm)262 in the presence of liquid culture medium lacking the osmoticum263 followed by further culture on agar-solidified medium for 30 days.264 Fresh weight of recovered calli was also measured as described ear-265 lier. Beside FW measurements at the start and end of culture period,266 the presence of viable, green callus tissue at the end of culture was267 monitored as well.268 SSP nuclease activity gels269 The activity of SSP nucleases (EC 3.1.30.1) was assayed on270 polyacrylamide gels, basically as described before (Jámbrik et al.,271 2011). In brief, PEG6000 treated calli were extracted with 10 mM272 Tris–HCl (Sigma–Aldrich, Budapest, Hungary), pH 8.0, 150 mM273 NaCl (Reanal, Budapest, Hungary), 14.6 mM 2-mercaptoethanol274 (Sigma–Aldrich) and 2% (w/v) polyvinyl-pyrrolidone (PVP, Merck,275 Darmstadt, Germany). Protein extracts (20 ␮g/well) were loaded276 on SDSand single-stranded DNA containing polyacrylamide gels277 along with a molecular weight marker (Sigma–Aldrich). Protein278 content of extracts was determined according to Bradford (1976). 279 After renaturation of enzymes, gels were incubated in Tris–HCl, pH280 6.8 (14 h, 39◦C) for assaying their activity, stained with 0.5 ␮g mL−1 281 ethidium bromide (Sigma–Aldrich, Budapest, Hungary) and exam-282 ined with an UV transilluminator. SSP nuclease activities appeared283 as clear bands, not stained with ethidium–bromide. Total nuclease284 activities on gels were quantified with the aid of CpAtlas®software285 and expressed as relative band intensities, where the value of con-286 trol activities was 1. The molecular weight of ssDNase isoenzymes287 was estimated with the UVI-TEC®software.288 Peroxidase activity gels289 PEG6000 treated calli were extracted at 4◦C with a buffer290 containing 100 mM KH2PO4/K2HPO4(VWR International Ltd.,291 Debrecen, Hungary), pH 7.2, 8 mM MgCl2(Reanal, Budapest,292 Hungary), 4 mM dithiothreitol (DTT, Sigma–Aldrich), 1% (v/v) Tri-293 ton X-100 (Reanal), 2% (w/v) PVP (Merck). After centrifugation (two 294 times for 30 min) at 15,000 × g with a Heraeus Biofuge, protein295 content of supernatants was assayed by the method of Bradford296 (1976). 10 ␮g protein was loaded onto each well of native 7.5%297 (w/v) polyacrylamide gels. Electrophoresis was performed at 4◦C,298 followed by gel staining for 30–60 min in a buffer containing299 100 mM sodium acetate (VWR), 10% (v/v) hydrogen peroxide and300 1 mM guaiacol. Peroxidase (E.C. 1.11.1.7) activity was visible due301 to dark-colored tetraguaiacol bands (Dixit et al., 2011). Guaiacol-302 peroxidase activities were quantified with the aid of CpAtlas® 303 software, and expressed as for SSP nucleases.304 Data analysis305 All experiments were performed at least four times with six306 parallel callus samples per experiment for each genotype and rep-307 resentative data are presented in the Results section. The mean ± SE308 of quantitative data was calculated and plotted with the aid of 309 Sigma Plot 10.0 software, where it was appropriate. Plots repre-310 sent mean ± SE values for different calli/the respective individual311 genotype. Quantitative data were subjected to statistical analysis 312 by two-way ANOVA. This involved All Pairwise Multiple Compari313 son Procedures (Holm-Sidak method) with an overall significance314 level of 0.05. This method allowed to analyze the effects of PEG con-315 centrations within a single genotype as well as differences between316 genotypes within a single PEG concentration. Differences were con317 sidered significant at P < 0.05. Fig. 1. Cluster dendrogram showing genetic distances between Quercus taxa involved in this study. Results 318 Genetic relationships between Quercus explants studied 319 Explants used in this study originated from six individual 320 mature trees from the Quercus (=Lepidobalanus) section co321 occurring at the Síkf˝ okút Research Site (Bükk Mts., Hungary). Leaf 322 morphological traits of the individual trees were studied in order 323 to estimate their taxonomical identity. Based on this, we identi324 fied three non-hybrid and three hybrid taxa (Table 1). In order 325 to estimate the genetic relatedness of selected trees a microsatel326 lite analysis was applied. Then genetic distances were calculated 327 between individuals based on the shared allele content of multilo328 cus genotypes derived from the six SSR loci analyzed (Fig. 1). This 329 revealed that A149 (Q. petraea) is genetically distinct to all other 330 individuals examined, including A75 (Q. pubescens) and B50 (Q. vir331 giliana). D89, a putative Q. petraea × Q. pubescens hybrid and A211, 332 a putative hybrid between Q. virgiliana and Q. polycarpa, appeared 333 to be in the same cluster with B50 and A75. D137, a putative hybrid 334 between Q. petraea and Q. dalechampii was genetically distant to 335 A149 and appeared to be in the same cluster, but at a relatively 336 high distance to all other individuals (Fig. 1). 337 The establishment and maintenance of tissue cultures from white 338 oak explants 339 We have induced and stabilized callus cultures for the first time 340 from six Quercus genotypes originating from Síkf˝ okút Research 341 Area (Table 1). A wide range of growth regulator concentrations 342 was tested (see Materials and methods section). Calli appeared after 343 30 ± 3 days of culture of young leaf explants. PGR combinations for 344 callus induction and maintenance as well as embryogenesis and 345 organogenesis were genotype dependent (data not shown). This is 346 a general rule for tissue cultures of related, but genetically different 347 plant taxa (see Duncan et al., 1985; Máthé et al., 2012 for exam348 ples). However, we have found a PGR combination, where all callus 349 samples originating from different oak genotypes were of similar 350 morphology (undifferentiated state), growth rate and viability. This 351 was WPM medium containing 4 mg L−1NAA and 0.5 mg L−1BA. At 352 this PGR content, growth of all stable cultures proved to be con353 stant since they were initiated (a time period of at least one year), 354 therefore they were suitable for osmotic stress experiments. 355 It is worth mentioning, that we could not establish an efficient 356 micropropagation system (i.e. mass plant regeneration without an 357 intermediary callus stage) from the aforementioned oak genotypes. 358 However the choose of PGRs (NAA and BA instead of 2,4-d and359 kinetin) excluded somaclonal variability during callus induction 360 Please cite this article in press as: Demeter Z, et al. Osmotic stress responses of individual white oak (Quercus section, Quercus subgenus) genotypes cultured in vitro. J Plant Physiol (2013), http://dx.doi.org/10.1016/j.jplph.2013.09.013 ARTICLE IN PRESS G Model JPLPH 51807 1–9 Z. Demeter et al. / Journal of Plant Physiology xxx (2013) xxx– xxx 5 Fig. 2. The effect of PEG concentration on fresh weight changes of calli with different taxonomic/genetic origin. Cultures were grown on WPM medium containing 4 mg L−1NAA and 0.5 mg L−1BA. (A) FW changes after PEG treatments; (B) FW changes during recovery of PEG treated calli. Mean ± SE values for different calli/the respective individual genotype are plotted. Asteriscs represent significant differences between PEG treatments for a single genotype, while lettercodes represent differences between callus lines within a single PEG treatment. Differences were considered to be significant at P < 0.05. and maintenance, thus callus cultures proved to be suitable for 361 osmotic stress experiments. 362 The effects of PEG6000 on water loss and recovery of Quercus363 callus cultures364 Two-way ANOVA revealed that there was a significant relation-365 ship (P < 0.05) between the genotype of individual oak trees and the366 effects of PEG treatments concerning all physiological parameters367 (growth, recovery and enzyme activities) studied (Figs. 2–5).368 The measurement of FW after treatment with the osmoticum369 revealed that for A149, significant water loss occurred at 10% 370 PEG6000 and increased progressively and significantly as PEG6000371 concentrations increased. In contrast, A75 and B50 were tolerant372 to osmotic treatment and significant water loss occurred only at373 40% PEG6000 (Fig. 2A). Putative hybrids were intermediate with 374 respect to the effects of the osmoticum: water loss occurred at 20%375 PEG6000 and increased at 40% PEG6000 (Fig. 2A). At recovery after376 osmotic stress, changes of FW showed that for A149, callus growth377 was slightly stimulated by 10% PEG6000, but inhibited significantly 378 at higher concentrations, while for A211 and B50 there was a tran-379 sient stimulation at 5–10% PEG6000 (Fig. 2B). After washout of the380 osmoticum there was a significant inhibition in the growth of A211 381 calli pretreated with 40% PEG6000 and pretreatment with 20–40% 382 PEG6000 inhibited the growth of B50 calli. Thus there was no recov383 ery in these cases. In case of culture lines D137, D89 and A75, there384 was a general inhibition of callus growth by PEG6000 (Fig. 2B).385 Statistical analysis of data revealed that PEG6000 induced water 386 loss of A75 calli was significantly different to all other callus lines.387 Pairwise comparison showed significant differences in PEG6000 388 induced water loss between A149, B50 and A211 only at higher 389 (20–40%) concentrations (Fig. 2A). Concerning callus growth dur390 ing recovery experiments, although significant differences were 391 observed between callus lines at certain PEG6000 concentrations 392 (Fig. 2B), significantly higher growth rate of A75 and B50 calli as 393 compared to lines A149, D137, D89 and A211 was not detected. 394 Concerning the presence of viable, compact, green callus tissue 395 during recovery experiments, we have made the following obser396 vations. For A149 (Q. petraea), this type of tissue was not present at 397 pretreatment with 10–40% PEG6000: it was replaced by necrotic398 like, browning tissue that suggested the accumulation of phenolic 399 compounds as a stress reaction (Fig. 3). In contrast, for A75 (Q. 400 pubescens) and B50 (Q. virgiliana), tissue viability persisted even 401 at 40% PEG (Fig. 3). D89 as a putative hybrid between Q. petraea 402 and Q. pubescens, was intermediate in this respect: callus viability 403 was lost only at 20–40% PEG. The putative hybrids D137 and A211 404 behaved similarly to A75 and B50 (Fig. 3). 405 Osmotic stress-related enzyme activities 406 Activity gels revealed that SSP nuclease isoenzyme(s) with 407 molecular weight(s) in the range of 45–55 kDa were present and 408 active in all culture lines. One dominant band of 50 kDa was 409 detected (Fig. 4A). In control cultures, this activity was weaker 410 in lines A149, D137 and D89 as compared to A211, B50 and A75 411 (Fig. 4A). The effects of PEG6000 treatments were dependent on 412 culture line/genotype. Except A75 (Q. pubescens) cultures, osmotic 413 stress induced transient increases in nuclease activities as com414 pared to controls, with maximal activities at 5–20% PEG6000. 415 Significant increases were detected for D89 (5–10% PEG6000), A149 416 (10–20% PEG6000) and A211 (20% PEG6000). In case of B50 (Q. vir417 giliana) there was only a slight stimulation of SSP nuclease activity 418 by PEG (Fig. 4A and B). In case of A75, PEG6000 did not increase 419 notably the enzyme activity, but at 5–10% PEG6000, two bands with 420 strong activities were detectable in the molecular weight range of 421 50 kDa (Fig. 4A and B). 5–20% PEG6000 induced the appearance of 422 one additional band with relatively weak activity in case of B50 and 423 A211 (Fig. 4A). 424 Two-way ANOVA revealed significant differences between SSP 425 nuclease activities of PEG6000 treated A149 and D137, B50, A75, 426 respectively. At 20% PEG, A149 was characterized by significantly 427 higher SSP nuclease activity than the rest of callus lines. Besides 428 A149, at 10% PEG6000, enzyme activity of A211 was significantly 429 higher as compared to A75 (Fig. 4B). 430 Concerning guaiacol peroxidase activities, one band with strong 431 activity appeared in all cultures. PEG6000 decreased this enzyme 432 activity in A149 (Q. petraea), although this decrease was not sig433 nificant (Fig. 5A and B). A non-significant decreasing effect was 434 observed in the case of putative hybrids D89 and A211 as well, 435 but transient increases were detectable at 10% and 5–10% PEG6000, 436 respectively (Fig. 5A and B). PEG treatments increased guaiacol per437 oxidase activities in the putative hybrid D137, Q. virgiliana (B50) 438 and Q. pubescens (A75) with peaks at 5% (D137, B50) and 20% 439 PEG6000 (A75) (Fig. 5A and B). An additional band of weak per440 oxidase activity was observed in case of D89, A211 and A75. For 441 A75, this band was present only after treatments with 20% PEG6000 442 (Fig. 5A). 443 Two-way ANOVA revealed significant differences between per444 oxidase activities of PEG6000 treated A149 and D137, B50, A75, 445 respectively (Fig. 5B). At 5% PEG6000, the enzyme activity of B50 446 was significantly higher, than A75. Concerning overall effects of the 447 osmoticum, peroxidase activities of the putative hybrid D89 did not 448 differ significantly to A149, but it had significantly lower activities, 449 Please cite this article in press as: Demeter Z, et al. Osmotic stress responses of individual white oak (Quercus section, Quercus subgenus) genotypes cultured in vitro. J Plant Physiol (2013), http://dx.doi.org/10.1016/j.jplph.2013.09.013 ARTICLE IN PRESS G Model JPLPH 51807 1–9 6Z. Demeter et al. / Journal of Plant Physiology xxx (2013) xxx– xxx Fig. 3. Representative recovery experiments with Quercus tissue cultures of different taxonomic/genetic origin. PEG treatments were followed by washout of the osmoticum and further culture on a medium supplemented with 4 mg L−1NAA and 0.5 mg L−1BA. Scale bar: 5 mm. than D137 and B50. A211 had significantly lower activities, than450 B50 and D137 (Fig. 5B). 451 Discussion452 We have established six novel stable callus lines from oak geno-453 types belonging to the Quercus section (=Lepidobalanus)of Quercus 454 subgenus. These cultures were suitable for comparing differences 455 in osmotic stress responses among genotypes. Based on previous456 work on the production of stable tissue cultures of Q. robur (Cuenca457 et al., 1999; Toribio et al., 2004), we established the proper culture458 media suitable for drought stress experiments with PGR content of459 4 mg L−1NAA and 0.5 mg L−1BA.460 Clear distinction from molecular markers between Q. petraea461 and Q. pubescens is often difficult (Salvini et al., 2009). However,462 microsatellite and isoenzyme data have previously demonstrated463 that Q. petraea and Q. pubescens are distinct species, but cross-464 ing between them is possible (Samuel et al., 1995; Bruschi et al.,465 2000). By applying 6 microsatellite loci, in this study the selected466 Q. petraea individual tree could be clearly distinguished from all 467 other Quercus (=Lepidobalanus) individuals, including Q. pubescens468 and putative hybrids (Fig. 1). Thus, significant genetic distances469 could be observed between individuals characterized by different470 leaf morphological traits. This is of particular importance, since dif471 ferent leaf morphologies of oaks do not necessarily reflect notable472 genetic differences (Curtu et al., 2007). The clear distinction of Q.473 petraea to all other individuals studied was confirmed by in vitro 474 morphogenesis experiments as well: this individual explant could 475 not regenerate roots, in contrast to all other explants studied, where 476 efficient root production was regularly observable (to be published477 elsewhere). Native gels revealed a single main peroxidase activity478 band for all culture lines. In case of D89, A211 and A75 a minor 479 additional band appeared. Interestingly, this band was inducible480 by PEG in case of A75, that is, it was detectable only during osmotic 481 stress (Fig. 5A, arrowheads). This additional band further supported 482 microsatellite data: in case of the genetically more distinct lines 483 A149 (Q. petraea) and D137 it was not present, while it appeared in 484 genetically related culture lines mentioned. In case of SSP nuclease 485 activity patterns, in the genetically related lines A211, B50 and A75, 486 two bands appeared at PEG6000 treatments, both in the molecu487 lar weight range of 45–55 kDa (Fig. 4). It should be noted however, 488 that many nucleases are glycoproteins and the presence or absence 489 of glycoside residues depends on the physiological state of cells 490 (Desai and Shankar, 2003). Therefore it is possible that double 491 bands detected reflect a single protein with different glycosylation 492 states. 493 Drought/osmotic stress, as a significant number of abiotic 494 stresses, leads to the increase of ROS in plant cells (Mittler, 2002). 495 Resistance to drought involves an increased capacity of scav496 enging ROS through superoxide dismutases (SOD), catalases and 497 peroxidases (Wang et al., 2003). Non-enzymatic and enzymatic 498 scavenging mechanisms are stimulated during summer midday 499 characterized by high light exposure, temperature and water defi500 ciency in Q. suber (Faria et al., 1996). In contrast, for drought and salt 501 sensitive Q. robur, among ROS scavenging systems, only superox502 ide dismutase (SOD) activity and isoenzyme pattern was modified 503 at exposure to NaCl (Sehmer et al., 1995). Drying of recalcitrant 504 Q. robur acorns is leading to the loss of embryo viability, associ505 ated with the accumulation of ROS and low levels of scavenging 506 enzymes (Hendry et al., 1992). In case of a drought tolerant Q. robur 507 genotype, elevated levels of ROS scavenging enzyme (SOD, ascor508 bate peroxidase, catalase, dehydroascorbate reductase, glutathione 509 reductase) activities were detected (Schwanz and Polle, 2001). In 510 the absence of a proper scavenging capacity, cellular structures and 511 macromolecules including DNA (single strand breaks) are signifi512 cantly damaged (Reddy et al., 2004). Following oxidative damage 513 Please cite this article in press as: Demeter Z, et al. Osmotic stress responses of individual white oak (Quercus section, Quercus subgenus) genotypes cultured in vitro. J Plant Physiol (2013), http://dx.doi.org/10.1016/j.jplph.2013.09.013 ARTICLE IN PRESS G Model JPLPH 51807 1–9 Z. Demeter et al. / Journal of Plant Physiology xxx (2013) xxx– xxx 7 Fig. 4. The effect of PEG on SSP DNase activities of Quercus calli of different genetic origin, grown on a medium supplemented with 4 mg L−1NAA and 0.5 mg L−1BA. (A) Representative gels. Arrows indicate the presence of additional bands; (B) enzyme activities represented by relative band intensities as measured with CP Atlas software. Mean ± SE values for different calli/the respective individual genotype are plotted. Asteriscs represent significant differences between PEG treatments for a single genotype, while lettercodes represent differences between callus lines within a single PEG treatment. Differences were considered to be significant at P < 0.05. of DNA, cells are characterized by the activity of nucleases involved514 in repair (Roldán-Arjona and Ariza, 2009).515 Several SSP nucleases inducible by drought/osmotic/salt stress516 have been identified. For example, barley BnucI is a salt stress517 inducible type I nuclease, a putative glycoprotein slightly smaller518 than 36 kDa from barley (Muramoto et al., 1999). Hydrogen per-519 oxide and drought stress induces the activity of several nuclease520 Fig. 5. The effect of PEG on guaiacol peroxidase activities of Quercus calli of different genetic origin, grown on a medium supplemented with 4 mg L−1NAA and 0.5 mg L−1 BA. (A) Representative gels. Arrowheads indicate the presence of additional bands. (B) Enzyme activities represented by relative band intensities as measured with CP Atlas software. Mean ± SE values for different calli/the respective individual genotype are plotted. Asteriscs represent significant differences between PEG treatments for a single genotype, while lettercodes represent differences between callus lines within a single PEG treatment. Differences were considered to be significant at P < 0.05. isoenzymes using ssDNA as substrate of molecular weights rang521 ing between 26 and 38 kDa in cauliflower seedlings. Two of them 522 appeared be the same enzyme, inducible by both stress factors 523 (Le´ sniewicz et al., 2010). The main nuclease isoenzyme proved to 524 be modulated by osmotic stress in this study appeared to have a525 molecular weight of 50 kDa (Fig. 4). To our best knowledge, nucle526 ases of similar molecular weight, with changing activity at abiotic 527 stresses are unusual in plants. It should be noted however, that a528 significant number of nucleases consist of subunits of lower size 529 (see Desai and Shankar, 2003 for a review). Thus, the nature of 530 Quercus nuclease of 50 kDa needs further investigation. 531 In case of drought sensitive individuals (A149, D89), PEG532 induced increase of SSP nuclease activity was accompanied by, 533 no significant changes or decreases of guaiacol-peroxidase activ534 ities in callus. In contrast, for drought-tolerant taxa (A75, B50), 535 PEG treatments led to unchanged or slightly increasing SSP nucle536 ase activities together with increases of peroxidase activities 537 (Figs. 4 and 5, see details below). This means that ROS cannot 538 Please cite this article in press as: Demeter Z, et al. Osmotic stress responses of individual white oak (Quercus section, Quercus subgenus) genotypes cultured in vitro. J Plant Physiol (2013), http://dx.doi.org/10.1016/j.jplph.2013.09.013 ARTICLE IN PRESS G Model JPLPH 51807 1–9 8Z. Demeter et al. / Journal of Plant Physiology xxx (2013) xxx– xxx Table 3 Overview of potential osmotic stress/drought tolerance of PEG6000 treated oak callus lines as shown by different physiological parameters. This classification was based on significant differences in physiological responses as revealed by two-way ANOVA. Physiological parameters FW Growth (recovery) Viability (recovery) ssDNase activity Guaiacol-peroxidase activity Potentially drought tolerant B50, A75 n.d. A211, B50, A75 B50, A75 D137, B50, A75 Potentially drought sensitive A149, D89 n.d. A149, D89 A149, D89, A211 A149, D89 Is the difference between A75 and B50 significant? Only at 40% PEG6000 Yes No No Yes n.d., not detectable on the basis of available data. be scavenged in drought-sensitive taxa, followed by DNA strand539 breaks leading to activity increases of nucleases probably involved540 in repair. Drought-tolerant taxa are expected to be able of scav-541 enging ROS efficiently, thus DNA strand breaks occur probably542 with less frequency. However, changes (increases or decreases) 543 of SSP nuclease activities were not proportional with changes of544 peroxidase activities for all genotypes and/or PEG concentrations545 (Figs. 4 and 5). This indicates that besides ROS levels, other mech546 anisms could modulate nuclease activities.547 In case of callus line A149 (Q. petraea), treatments with PEG6000548 induced concentration-dependent water loss, as seen by the549 decrease of callus fresh weight. The capacity of calli to grow and 550 to produce green, viable tissues decreased after removal of 20–40%551 and 10–40% PEG6000 respectively, as shown by recovery exper-552 iments (Figs. 2 and 3). Meanwhile, peroxidase activity decreased 553 and SSP nuclease activity increased (Figs. 4 and 5). All these effects554 of osmotic stress in a model experiment (Table 3) confirm pre-555 vious findings (Thomas et al., 2002) and our hypothesis, that is,556 sessile oak is generally a drought/osmotic stress sensitive species. 557 In contrast, cultures of A75 (Q. pubescens) and B50 (Q. virgiliana)558 are resistant to PEG6000 induced water loss and they are able559 to recover even after treatments with high concentrations of the560 osmoticum (even though their increase in fresh weight is inhib561 ited by PEG in recovery experiments, Figs. 2 and 3). Osmotic stress562 increases peroxidase activities and induces only slight increases of 563 SSP nuclease activities in these cultures. Overall, these results con-564 firm previous findings and our hypothesis, that is, tissue cultures565 of Q. pubescens and Q. virgiliana origin are drought/osmotic stress 566 tolerant (Table 3). Moreover, microsatellite data show a close relat567 edness between A75 and B50 (Fig. 1). For D89, PEG6000 induced568 water loss was intermediary as compared to A149 and A75, as569 shown by changes of callus fresh weight and the persistence of 570 callus recovery after treatment with 10% PEG (Figs. 2A and 3). On571 the other hand, osmotic stress induced the increase of ssDNase572 activity and except treatments with 10% PEG, the decrease of per573 oxidase activity (Figs. 4 and 5). Overall, these parameters suggest 574 that drought response of D89 is intermediary between A149 and575 A75, adding further proof to the idea that D89 is a hybrid between576 Q. petraea and Q. pubescens. However, all physiological parameters 577 suggest that cultures of D89 are more tolerant to drought, than578 A149, but still can be considered as drought sensitive (Table 3).579 Peroxidase activities of PEG6000 treated D137 calli increased as of 580 drought resistant genotypes (Fig. 5 and Table 3), even though we581 have initially hypothesized that as a hybrid between Q. petraea and582 Q. dalechampii, this genotype is relatively drought sensitive. Indeed,583 microsatellite data confirmed that D137 is genetically closer to 584 the cluster containing Q. pubescens, than to Q. petraea (Fig. 1). On585 the other hand, Q. dalechampii lineage could also explain a higher586 drought tolerance. Concerning A211, its water loss and ability to587 recover was intermediary between drought sensitive A149 and 588 tolerant A75 (Figs. 2A, 3 and Table 3), but PEG6000 induced SSP589 nuclease and peroxidase activity changes were similar to A149 590 (Figs. 4 and 5). Morphological and microsatellite data were not com591 parable for this culture line. Leaf morphology traits data suggested 592 that A211 is a hybrid between Q. virgiliana and Q. polycarpa, but593 the analysis of SSR markers suggested its close relatedness to the 594 analyzed Q. pubescens tree (Table 1 and Fig. 1). Thus, even though 595 this individual tree is putative hybrid between a drought sensitive 596 and a drought resistant oak taxon, genes/proteins responsible for 597 drought tolerance seem to be weakly expressed. 598 It should be noted that a given oak species can be character599 ized by high intra-populational genetic variability (Herzog, 1996)600 and as a consequence, differences between physiological responses 601 of individuals may occur. Thus, intra-populational spectrum of 602 drought tolerance of oaks in the Síkf˝ okút LTER area needs further 603 studies. 604 Table 3 shows the estimated osmotic stress/drought tolerance of 605 the six oak culture lines studied as shown by significant physiologi606 cal responses to PEG treatments. All physiological changes induced 607 by osmotic stress indicated that drought sensitivity of Q. petraea, Q. 608 pubescens, and Q. virgiliana cultures confirmed previous laboratory 609 and field studies including ecological requirements of their pop610 ulations (Cochard et al., 1992; Damesin and Rambal, 1995; Gallé 611 et al., 2007; Siam et al., 2009; Rodríguez-Calcerrada et al., 2010; 612 Ofletea et al., 2011) and proved our hypothesis: such in vitro cul613 tures can be additional model systems for studying stress responses 614 of taxa with unknown drought responses (D89, D137 and A211 615 in the present study). Interestingly, B50 (Q. virgiliana) and A75 (Q. 616 pubescens) had significantly different responses in recovery exper617 iments and concerning peroxidase activities (Figs. 2, 3 and 5 and 618 Table 3), even though our data and previous findings suggest that 619 both Q. pubescens and Q. virgiliana are drought tolerant species (see 620 Table 3; Siam et al., 2009; Ofletea et al., 2011 for example). In gen621 eral, microsatellite data confirmed relatively high genetic distances 622 between drought sensitive and drought tolerant oak individuals. 623 The present work offers an in vitro model of studying physiologi624 cal responses of oak to drought, by using tissue cultures established 625 from selected individual trees with different taxonomic identity co626 occurring in the same forest stand. Further studies both in the field 627 and laboratory, the latter concerning tissue cultures from different 628 individuals from the same population will reveal natural variability 629 in drought tolerance of white oaks. We suggest that in vitro callus 630 cultures as simplified experimental systems are helpful tools con631 tributing to the modeling of drought tolerance of field grown oak 632 plants with different genetic background and could be of general 633 applicability for plant stress biology research. 634 Acknowledgements 635 The study was supported by National Research Foundation 636 (OTKA No. K68397 and K101552) and European Union and the 637 European Social Fund co-financed project TÁMOP-4.2.2/B-10/1638 2010-0024. 639 References 640 Aas G. Morphologische und ökologische Variation mitteleuropäischer Quercus641 Arten: Ein Beitrag zum Verständnis der Biodiversität. Libri Botanici: Band 19. 642 Eching: IHW-Verlag; 1998. p. 213. 643